feat: Implement distributed pool with consistent hashing

[quasi]

This commit introduces a distributed pool type (distributed-pool) to the pooler library, utilizing consistent hashing to distribute items across multiple underlying pool instances (nodes).

Key features and changes:

1. distributed-pool Structure:

   • Manages a collection of named nodes, where each node is a standard pool instance.
   • Uses a consistent hash ring to map keys to specific nodes.
   • Includes configurable replica counts for nodes on the hash ring to improve key distribution.

2. Consistent Hashing Implementation:

   • hash-ring-node structure to represent nodes on the ring.
   • Functions to add (add-node-to-ring), remove (remove-node-from-ring), and retrieve (get-node-id-for-key-from-ring) nodes based on hashed keys.
   • Uses sxhash for string hashing.

3. Core Distributed Pool API:

   • make-distributed-pool: Constructor to create a distributed pool with specified node configurations (name, item-maker, capacity, etc.).
   • fetch-from-distributed-pool: Fetches an item from the appropriate node based on a provided key.
   • return-to-distributed-pool: Returns an item to the appropriate node based on a key.

4. Node Management:

   • add-node-to-distributed-pool: Dynamically adds a new node (pool instance) to an existing distributed pool.
   • remove-node-from-distributed-pool: Removes a node from a distributed pool. (Note: Item draining/migration is not currently implemented for removed nodes).

5. Utility and Export:

   • get-node-for-key: Retrieves the actual pool object for a given key (used internally and available for advanced use).
   • All new user-facing functions and structures are exported in package.lisp.

6. Testing:

   • A new test suite (tests/test-pooler.lisp) has been added.
   • Tests cover basic distributed pool operations, node addition/removal, key distribution characteristics, and error handling.
   • pooler.asd has been updated to define and run these tests via (asdf:test-system :pooler/tests).

This implementation provides a foundational distributed pooling mechanism. Future enhancements could include more sophisticated item draining on node removal, dynamic adjustment of replicas, or different hashing strategies.